DE69122956T2 - METHOD AND COMPOSITIONS FOR INHIBITING ANGIOGENESIS - Google Patents

Abstract

Pharmaceutical compositions for delivering an effective dose of an angiogenesis inhibitor consisting of a specific inhibitor of collagenase, for inhibition of blood vessel ingrowth and tube formation in vivo and capillary endothelial cell proliferation and migration in vitro, which is non-cytotoxic to normal cells. The compositions are delivered topically or locally using implants or injection. Examples of specific collagenase inhibitors include purified cartilage-derived inhibitor, collagenase binding peptides and antibodies which block the active site or metal binding regions of collagenase, and compounds which inhibit enzymes activating collagenase, such as plasmin inhibitor.

Description

Background of the invention

The United States Government has rights to this invention pursuant to a grant from the Federal Health Authorities.

Angiogenesis, the process by which new capillaries form, plays an important role in numerous physiological events, both normal and pathological. Angiogenesis is a complex process characterized by the infiltration of the basement membrane by capillary endothelial cells, the migration of cells through the matrix toward a stimulus, and subsequent cell proliferation leading to the formation of new capillary tubes.

Although several proteases play key roles in angiogenesis, it has not yet been shown that one enzyme is crucial for all three components of angiogenesis, namely the migration of endothelial cells and the proliferation and ingrowth of capillary tubes. See, for example, Mignatti et al., Cell 47, 487-489 (1986) and Rifkin et al., Acta. Biol. Med. Germ. 40, 1259-1263 (1981), who are of the opinion are that a number of enzymes in a proteolytic cascade, including plasminogen activator and collagenase, must be inhibited to inhibit angiogenesis. The plasminogen activator and collagenase interact via plasmin, which is cleaved from plasminogen by the plasminogen activator and cleaves the latent form of collagenase to yield the active form. As a result, it has not been possible to show that an enzyme inhibitor is capable of inhibiting angiogenesis without cytotoxicity.

Under normal circumstances, angiogenesis is associated with events such as wound healing, corpus luteum formation, and embryonic development, as discussed by Folkman et al., Science 243, 1490-1493 (1989). However, some serious diseases are also dominated by abnormal neovascularization, such as the growth of solid tumors and metastases, some types of eye disease, and rheumatoid arthritis. For a review, see Auerbach et al., J. Microvasc. Res. 29, 401-411 (1985); Folkman, Advances in Cancer Research, eds. Klein and Weinhouse, pp. 175-203 (Academic Press, New York 1985); Patz, Am. J. Ophthalmol. 94, 715-743 (1982) and Folkman et al., Sciences 221, 719-725 (1983). For example, there are a number of eye diseases, many of which result in blindness, in which ocular neovascularization occurs as a response to the disease. These eye diseases include diabetic retinopathy, neovascular glaucoma, inflammatory diseases, and eye tumors (e.g., retinoblastoma). In addition, there are a number of other eye diseases that are also associated with neovascularization, such as retrolental fibroplasia, uveitis, about twenty Eye diseases associated with choroidal neovascularization and about 50 eye diseases associated with neovascularization of the iris.

Current treatment of these diseases is inadequate, especially when neovascularization has already occurred, which often results in blindness. Studies have led to the view that vasoinhibitory factors present in normal ocular tissue (cornea and vitreous) are lost in the diseased state. The potential therapeutic benefit that a naturally occurring inhibitor of angiogenesis could have in combating diseases in which neovascularization plays a crucial role has led to a long-standing search for angiogenesis inhibitors. An angiogenesis inhibitor could play an important therapeutic role in alleviating the course of these diseases, as well as being a useful tool in studying their etiology.

A number of investigators have reported that extracts from cartilage, one of the few avascular tissues in the body, can inhibit angiogenesis: Eisenstein et al., Am. J. Pathol. 81, 1-9 (1987); Pauli et al., J. Natl. Cancer Inst. 67, 55-74 (1981); Brem and Folkman, J. Exp. Med. 141, 427-439 (1975); Langer et al., Science 193, 70-72 (1976); Langer et al., Proc. Natl. Acad. Sci. USA 77, 432-435 (1980) and Lee and Langer, Science 221, 1185-1187 (1983). Langer et al. have shown that cartilage extracts containing a collagenase inhibitor delay tumor-induced and inflammation-induced neovascularization of the cornea and conjunctiva when administered either by infusion or by prolonged release from a polymeric implant.

Other inhibitors that inhibit collagenase in vitro do not inhibit angiogenesis when tested in vivo, such as α2-macroglobulin and tissue inhibitory metalloproteinase (TIMP). Angiogenesis has several major components, such as capillary endothelial cell proliferation, endothelial cell migration, and collagenase formation. To inhibit angiogenesis, a compound must inhibit capillary endothelial cell proliferation, endothelial cell migration, and collagenase, either alone or in combination with one or more other factors. At the doses tested, TIMP inhibits capillary endothelial cell proliferation but not angiogenesis, and protamine inhibits vessel ingrowth but not capillary endothelial cell proliferation. Protamine is cytotoxic at concentrations at which angiogenesis appears to be inhibited.

Extracts from various other tissues have also been shown to have anti-angiogenesis activity, see for example the review by D'Amore, Prog. Clin. Biol. Res. 221, 269-283 (1986). Although various molecules have been identified which inhibit various aspects of angiogenesis, such as cell proliferation or cell migration, no single molecule capable of inhibiting angiogenesis in vivo has been identified in the prior art until applicants' report in Science 248, 1408-1410 (1990), namely as described in US application No. 07/027,364, filed March 18, 1987, Langer et al. This application, which is a continuation-in-part earlier applications reporting a partially purified protein extract isolated from cartilage and having anti-angiogenesis activity, discloses the purification of a single protein which inhibits collagenase and which is further believed to be an inhibitor of angiogenesis, tumor-induced vascularization, connective tissue degradation and tumor cell invasion and which is therefore useful for the treatment of corneal ulcers and rheumatoid arthritis and psoriasis, namely due to the inhibition of collagenase.

An object of the present invention is to provide pharmaceutical compositions and methods of using them for treating diseases associated with abnormal angiogenesis.

Another object of the present invention is to provide pharmaceutical compositions and methods of using them for inhibiting the proliferation and migration of capillary endothelial cells.

Another object of the present invention is to provide pharmaceutical compositions and methods of use thereof which inhibit the proliferation and migration of capillary endothelial cells and collagenase or the activation of collagenase in pharmaceutically acceptable dosages.

Another object of the present invention is to provide controlled release pharmaceutical compositions for topical application and methods of using them to inhibit angiogenesis.

Summary of the invention

An isolated and purified protein, wherein the protein: a) is a normal component of cartilage, b) has an apparent molecular weight by SDS-PAGE under reducing conditions of about 34 kD when isolated without exposure to acidic pH or of about 28 kD after exposure to pH 2.0, c) is a collagenase inhibitor, and d) produces an avascular zone in a chorioallantoic membrane assay when used in an amount of about 4 µg.

The protein can be used to inhibit blood vessel growth in vivo and capillary endothelial cell proliferation and migration in vivo. The protein can be extracted from bovine scapular cartilage using 2 M NaCl followed by a series of precipitation and column chromatography steps, or it can be expressed from nucleotide sequences encoding the protein and isolated using probes based on the NH2-terminal protein sequence.

The effective dosage for inhibition of angiogenesis in vivo, defined as inhibition of capillary endothelial cell proliferation and migration and blood vessel ingrowth, is extrapolated from in vitro inhibition assays. Effective dosages in vitro range from 1 nM, namely for inhibition of capillary endothelial cell migration, to 100 nM, namely for inhibition of capillary endothelial cell proliferation. The preferred dosage in vivo is between 16 nM and 50 to 75 nM. The effective dosage depends on the The method and means of delivery, which may be localized or systemic. For example, in some applications, such as the treatment of psoriasis or diabetic retinopathy, the inhibitor may be delivered in a topical ophthalmic carrier. In other applications, such as the treatment of solid tumors, the inhibitor is delivered using a biodegradable polymeric implant.

In vitro assays have been developed to screen for compounds that are effective as inhibitors of angiogenesis in vivo. These assays represent a major event required for angiogenesis, so that inhibition in the assays is representative of inhibition of angiogenesis in vivo. The events studied include proteolytic degradation of the extracellular matrix and/or basement membrane, proliferation of endothelial cells, migration of endothelial cells, and formation of capillary tubes.

Detailed description of the invention

It has now been found that, contrary to numerous reports in the literature, all of the various important factors of angiogenesis, namely the proliferation and migration of endothelial cells, the formation of tubes and the ingrowth of blood vessels, can be inhibited by specifically inhibiting only collagenase. This is extremely surprising in view of the many different reaction pathways and enzymes involved in angiogenesis. It was not expected to find a common element that allows sufficiently significant specific inhibition of angiogenesis without cytotoxicity.

In vitro assays to determine the potency and effective concentration of angiogenesis inhibitors in vivo

To screen for compounds that inhibit collagenase activity, proteolytic degradation of the extracellular matrix or basement membrane, migration of endothelial cells, and capillary tube formation, assays were used to determine (1) potential efficacy in vivo as inhibitors of angiogenesis and (2) effective concentration. As discussed above, all of these elements of angiogenesis must be inhibited in vivo for a composition to be effective. Collagenase inhibitors have been found to be effective in all of these assays when provided in an appropriate carrier and at an appropriate concentration. The assays are therefore used first to determine whether the compound is a collagenase inhibitor and then whether it is effective in the other assays and at what dosage.

The essential elements of these assays include the use of endothelial cells as the target of the assay and the stimulation of the endothelial cells with known angiogenesis factors in order to compare the effect of a putative inhibitor in unstimulated endothelial cells with that in stimulated ones. The endothelial cells that can be used include capillary endothelial cells and umbilical vein endothelial cells. Factors that can be used to stimulate the endothelial cells include acidic or basic fibroblast growth factor.

The following assays are based on assays described in the prior art that are used to screen for Collagenase inhibitors were modified to inhibit angiogenesis in vivo.

Assay for inhibition of collagenase activity.

One unit of collagenase inhibitory activity is defined as the amount of protein required to inhibit one unit of corneal collagenase by 50%. One unit of collagenase causes 10% cleavage of collagen in 2.5 hours at 37°C. Collagenase activity is determined using an assay developed by Johnson-Wint, Anal. Biochem. 68, 70 (1980). Type I collagen is purified from rat tail tendon by solubilization in 0.5 M acetic acid and repeated precipitation with sodium phosphate and then acetylated using [14C]-acetic anhydride. The radiolabeled collagen is dissolved in 0.5 M acetic acid together with sufficient unlabeled purified type I collagen to give a solution of 2 mg/ml with a specific activity of approximately 350,000 cpm/ml. The collagen solution is dialyzed twice at 4ºC against 0.15 M sodium phosphate buffer, pH 7.6. Then 96-well tissue culture plates are set up at 45º and aliquots of the dialyzed solution (25 µl, 8,750 cpm) are added to the lower half of each well. Incubating the plates at 37ºC for 1 hour allows fibrils to form. The plates are then immersed in water for 1 hour and then allowed to dry overnight.

Collagenase is obtained from the culture medium of explanted bovine corneae. Corneae are excised from bovine eyes and incubated with clinidine solution and then with Hank's balanced salt solution containing 2% penicillin-streptomycin. They are then cut into 4 mm2 pieces and placed in T150 tissue culture flasks containing Dulbecco's modified Eagle's medium, 5% fetal bovine serum and 2% penicillin-streptomycin solution. Then 0.5 ml of cytochalasin B (1 mg/ml in dimethyl sulfoxide) is added to 100 ml of the medium. One flask contains approximately 10 corneas/50 ml of medium. The flasks are incubated at 35ºC in an incubator in an atmosphere of 95% humid air and 5% CO₂. The medium is collected 7 days after the corneas have been explanted and stored at 4ºC. Fresh medium is then added to the corneas and a second collection is made on day 14. The collagenase-containing medium is stored under sterile conditions at 4ºC for up to 2 months. Before use, latent collagenase contained in the culture medium is activated by trypsin treatment by incubating the medium for 7 min at 37ºC with one-tenth volume of trypsin at a concentration of 1 mg/ml. After activation, one-tenth volume of soybean trypsin inhibitor (5 mg/ml) is added to inactivate the trypsin. The enzyme is then diluted with collagenase assay buffer (CAB) (0.05 Tris-HCl, pH 7.6, containing 0.2 M NaCl, 0.001 M CaCl₂ and 0.01% sodium azide, CAB) so that 100 µl of collagenase solution cleaves collagen (5 µg) to 10% in 2.5 h at 37ºC.

The samples to be tested for collagenase inhibitory activity are dissolved in CAB or dialyzed against it. 100 µl of the samples and 100 µl of the diluted enzyme solution are mixed together, added to the collagen-containing wells and incubated for 2.5 h at 37ºC. The supernatants, which contain soluble radiolabeled degradation products, are counted in a Beckman Model LS 1800 scintillation spectrometer. Assay for the inhibition of capillary endothelial cell proliferation.

Proliferation of capillary ECs in response to an angiogenic stimulus is a critical component of neovascularization, as discussed by Ausprunk and Folkman, J. Microvasc. Res. 14, 153-65 (1977). By using the specific cells involved in angiogenesis and stimulating them with known angiogenesis factors, in this case acidic fibroblast growth factor (aFGF), as reported by D'Amore and Klagsbrun, J. Cell Biol. 99, 1545-1549 (1984), it is possible to mimic the process of angiogenesis in vitro. This type of assay is the assay of choice to demonstrate stimulation of capillary EC proliferation by various angiogenic factors, see the review by Shing et al., Science 223, 1296-1298 (1984).

The number of endothelial cells in the culture can be rapidly determined based on the colorimetric determination of cellular acid phosphatase activity, as described by Connolly et al., J. Anal. Biochem. 152, 136-140 (1986). Capillary endothelial cells are plated on day 1 in gelatin-coated 96-well tissue culture dishes (2 x 10³/0.2 ml). On day 2, cells are returned to Dulbecco's modified Eagle's medium (Gibco) with 5% calf serum (Hyclone) (DMEM/5) and aFGF (2 ng/ml) (FGF Co.) and increasing concentrations of compound are added. Compound is added in volumes not exceeding 10% of the final volume. Control wells containing only phosphate buffered saline (PBS) (Gibco) and PBS + aFGF are provided. On day 5, the Media is removed and cells are washed with PBS and then lysed in 100 µl of buffer containing 0.1 M sodium acetate (pH 5.5), 0.1% Triton X-100 and 100 mM p-nitrophenyl phosphate (Sigma-104 phosphatase substrate). After incubation for 2 hours at 37ºC, the reaction is stopped by adding 10 µl of 1 N NaOH. Color development is determined using a fast microplate reader (Bio-Tek) at 405 nm.

According to Connolly et al., there is a linear relationship between acid phosphatase activity and the number of endothelial cells up to 10,000 cells/well. In addition, standard curves for acid phosphatase activity are constructed using known cell numbers to ensure that enzyme levels reflect the true EC number. Percent inhibition is determined by comparing the number of cells from wells exposed to the stimulus with those exposed to both stimulus and inhibitor. Each point represents the mean of 4 control wells and 3 inhibitor wells, with the SEM of the mean being less than 7%. The accuracy of the data is confirmed by electronic cell counting assays.

The effect of the compound on DNA synthesis of capillary ECs can be further determined by the suppression of radioactive thymidine incorporation into capillary ECs in response to aFGF.

Assay for inhibition of capillary endothelial cell migration.

To investigate the effects of CDI on the migration of capillary ECs, a modification of the chamber technique by Boyden, Falk et al., J. Immunol. 118, 239-247 (1980). A blank-well Boyden chamber as described in J. Exp. Med. 115, 453-456 (1962) consists of 2 wells (upper and lower) separated by a porous membrane. A known concentration of growth factor is added to the lower wells and a predetermined number of cells and the compound of interest are added to the upper wells. The cells bind to the upper surface of the membrane, migrate across, and bind to the lower membrane surface. The membrane can then be fixed and stained for counting using the method of Glaser et al., Nature 288, 483-484 (1980).

Migration is determined using blank well chambers (Neuroprobe, #025-187) and 8 µm pore polycarbonate membranes (Nucleopore) precoated with fibronectin (6.67 µm/ml in PBS) (human, Cooper). Basic FGF (Takeda Co.) diluted in DMEM with 1% calf serum (DMEM/L) is added to the lower well at a concentration of 10 ng/ml. 5 x 10⁵ capillary ECs/ml and increasing concentrations of the compound of interest are added to the upper wells. DMEM/L is added to the control wells, either with or without bFGF. Migration chambers are incubated for 4 h at 37°C in 10% CO₂. Then the cells on the upper surface of the membrane are wiped by pulling the membrane over a wiper blade (Neuroprobe). The cells that have migrated through the membrane to the lower surface are fixed in 2% glutaraldehyde and then with methanol (4ºC) and stained with hematoxylin. Migration is quantified by counting the cells on the lower surface in 16 oil immersion fields (OIF). Each point represents the mean +/- SEM of 4 wells. In the control wells without bFGF, the number of migrated capillary ECs is 61 +/- 7 cells.

Assay for inhibition of neovascularization in vitro.

The chicken chorioallantoic membrane assay (CAM) described by Taylor and Folkman, Nature (London) 297, 307-312 (1982) is used to determine whether the compound can inhibit neovascularization in vivo. The effect of the compound on growing embryonic vessels is studied using chicken embryos in which capillaries appear in the yolk sac after 48 hours and grow rapidly over the next 6 to 8 days.

On the third day of development, the fertilized chicken embryos are removed from their shells and placed in plastic petri dishes (1005, Falcon). They are kept at 37ºC in a humidified atmosphere with 5% CO2. On the sixth day, 4 µg samples of the purified CDI are mixed in methylcellulose disks and placed on the surfaces of the growing CAMs above the dense subectodermal plexus. After the CAMs have been exposed to CDI for 48 h, the avascular zones, free of capillaries and small vessels, are examined using a binocular dissecting microscope at x7-10 magnification. The CAMs are then injected intravascularly with Indian ink/liposyn as previously described by Taylor and Folkman (1982).

Tissue samples are fixed in formalin at room temperature and rinsed with 0.1 M cacodylate buffer, pH 7.4. Then samples are embedded in JB-4 plastic (Polysciences) at 4ºC and sectioned using a microtome Reichert 2050 sections of 3 µm are prepared. The sections are stained with toluidine blue. The micrographs are taken with a Zeiss photomicroscope using Kodak TM x100 and a green filter.

Assay for inhibition of capillary tube formation.

Capillary tube formation is an important component of angiogenesis. An assay that indicates capillary tube formation activity consists of applying the compound to or between collagen gels and determining whether capillary-like tubes are formed by cultured BCE cells. In the control collagen gels that were not treated with the inhibitor, a monolayer of BCE cells results in the form of a network of branched, anastomosing strands of endothelial cells that resemble capillary beds in vivo. In contrast, collagen gels treated with an inhibitor show severe disruption of BCE cell reorganization, so that organization into capillary-like tubes appears to be prevented.

Examples of in vitro assay-based determination of the anti-angiogenic activity of the cartilage-derived inhibitor in vivo.

Purified cartilage-derived protein (CDI) was isolated. It is characterized by having at least three different activities that are important for the inhibition of angiogenesis. Due to these activities, the protein can be administered in a pharmaceutically acceptable carrier at a dosage that stimulates angiogenesis in vivo by local inhibition of the proliferation and migration of capillary endothelial cells and the Ingrowth of blood vessels is inhibited.

The term CDI as used herein refers to any protein that specifically inhibits eukaryotic collagenase that is encoded by a nucleotide sequence that hybridizes under standard conditions to the nucleotide sequence encoding the inhibitor protein isolated from cartilage, as described below, or from chondrocyte or fibroblast culture fluid. Since CDI is not species specific, it is possible to use a CDI isolated from a non-human source to treat humans. CDI-like proteins that specifically inhibit collagenase are produced by a number of different tissues. Therefore, CDI proteins can also be expressed from genetically engineered sequences isolated by standard techniques from a library prepared from chondrocytes or other cells, e.g., fibroblasts capable of differentiating into cartilage, or from other mammalian cells, using as probes labeled synthetic oligonucleotides encoding portions of the CDI isolated from the cartilage. These probes can be synthesized using commercially available equipment or by a contract synthesis company. Methods for screening libraries using labeled probes or antibodies to a purified protein are well known to those skilled in the art, see, for example, Maniatis et al., Molecular Cloning: A Laboratory Manual (Cold Spring Harbor, NY 1989).

The partial purification and characterization of the protein with collagenase inhibiting activity are detailed in US Application No. 07/027,364, filed on March 18, 1987. The cartilage-derived inhibitor (CDI) is isolated from bovine cartilage by extraction with 2 M NaCl, followed by precipitation with hydrochloric acid and two precipitations with ammonium sulfate (25%, 60%). The inhibitor is applied to an A-1.5m gel filtration column. Further purification is achieved by ion exchange chromatography, a second gel filtration step and reversed-phase high performance liquid chromatography.

CDI can be isolated from cartilage. Calf shoulder blades are obtained from a slaughterhouse within 24 hours of slaughter. The half of the cartilage, furthest from the bone, is excised and scraped first with a periosteal scraper (Arista) and then with a scalpel blade (No. 10, Bard-Parker) until free of connective tissue. The cartilage strips are stored at -20ºC until use.

Frozen cartilage is thawed and minced in a Cuisinart Model DLC 10 into 1/8 in3 pieces. Minced strips (1 kg per batch) are extracted in 8 L of 0.01 M 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, pH 7.6, containing 2 M NaCl, 3 mM EDTA, and 0.02% sodium azide at 4ºC for 48 h. The extract is filtered through a Nitex filter (153 µm mesh, Tetko), centrifuged at 12,500 xg for 1 h at 4ºC, and then filtered through both Nitex and several layers of gauze. The filtered extract, approximately 7.4 l, is concentrated using a Pellicon cassette system (Millipore) equipped with a PTCG cassette (molecular weight cutoff, Mr 10,000). The concentrate is reprecipitated against 10 mM Tris-HCl, pH 7.6, containing 0.15% NaCl and 0.02% sodium azide. contains, dialyzed and stored at -20ºC.

To acid precipitate inhibitor impurities, 1 N HCl is added dropwise to the concentrated extract until the pH reaches 2.0. The extract is then centrifuged at 25,000 x g for 30 min and the pellet is discarded. 1 N NaOH is then added dropwise to the supernatant until the pH reaches 7.6. The extract is then centrifuged at 25,000 x g for 30 min and the pellet is discarded to remove any acid-insoluble impurities from the inhibitor-containing solution.

Then (NH₄)₂SO₄ is added to the extract to 25% saturation and the precipitate is discarded after centrifugation. Then more (NH₄)₂SO₄ is added to 60% saturation. The 60% saturated solution is centrifuged and the inhibitor-containing pellet is stored at -20ºC.

The thawed pellet is dissolved in 0.01 M Tris-HCl, pH 7.6, containing 4 M guanidine-HCl, 0.001 M CaCl2, and 0.2% sodium azide and applied to an A1.5m Sepharose size exclusion column (Pharmacia, Piscataway, NJ) (2.5 x 90 cm) at 4°C at a flow rate of 30 ml/h. Effluent from the column is monitored spectrophotometrically at 280 or 220 nm. The inhibitor elutes at an apparent molecular weight of 28,000 Da. All fractions containing inhibitor after the assay for collagenase activity are pooled and concentrated in an Amicon stirred cell equipped with a YM10 membrane. The retained material is dialyzed against 0.05 M Tris-HCl, pH 7.6, containing 0.05 M CaCl₂ and 0.05% sodium azide and then applied to a Bio-Rex 70 cation exchange column (200-400 mesh, 2.6 x 7.5 cm). The column is then washed with 10 column volumes of 0.05 M Tris-HCl, pH 7.6, containing 0.05 M CaCl2 and 0.05% sodium azide to elute non-absorbing protein. The bound material is eluted with a salt gradient, namely 0.05 to 0.75 M NaCl in 0.05 M Tris-HCl, pH 7.6, containing 0.001 M CaCl2 and 0.01% sodium azide. The inhibitor binds to the column at low salt concentrations, namely 0.05 M NaCl, and elutes up to about 0.2 M NaCl.

Fractions containing the inhibitor are concentrated in the stirred cell and dialyzed against collagenase assay buffer, CAB, 0.05 M Tris-HCl, pH 7.6, containing 0.2 M NaCl, 0.001 M CaCl2, and 0.01% sodium azide. The dialysate is applied to a column of Sephadex G-75 (superfine, Pharmacia) (1.6 x 50 cm) equilibrated with CAB at a flow rate of 10 ml/h. Fractions containing the inhibitor elute after the dominant peak of a material absorbing at 280 nm and are pooled and concentrated in the stirred cell to a volume of 1 ml. An equal volume of 0.05% trifluoroacetic acid in water is then added to the sample. Reverse phase high performance liquid chromatography is then performed on a C4 column. A linear gradient from 0.05% trifluoroacetic acid in water to 0.05% trifluoroacetic acid in acetonitrile is used to elute the bound protein. Aliquots of 100 µl are removed and immediately dialyzed against CAB to perform the assay for collagenase activity.

The collagenase inhibitor is purified to apparent homogeneity by reversed-phase high performance liquid chromatography. Total protein, total units, The specific activity obtained and the purification after each step are given in Table I. Table I: Purification of a bovine cartilage-derived collagenase inhibitor

a One unit is defined as the amount of protein required to inhibit one unit of corneal collagenase by 50%. One unit of collagenase causes 10% cleavage of collagen in 2.5 hours at 37ºC.

A CDI-like protein was isolated from conditioned media (HAM serum-free F12 medium supplemented with ascorbic acid, Gibco) in which bovine scapula chondrocytes had been cultured by extensive dialyzing of the medium against distilled water and subsequent application of the dialysate to a Sephadex G75 gel filtration column (superfine, Pharmacia). (1.6 x 50 cm) in 0.05 M Tris-HCl, pH 7.6, containing 0.2 M NaCl, 0.001 M CaCl2, and 0.01% sodium azide at a flow rate of 10 ml/h. Fractions were collected and gave a single band of molecular weight approximately 34,000 on SDS-PAGE under reducing conditions. These fractions inhibited collagenase as determined above. The fractions were also active in both the proliferation and migration assays and in the CAM assay for inhibition of angiogenesis, giving 80% inhibition at 4 µg.

Automated Edman degradation of CDI was performed using an Applied Biosystems, Inc. Model 477A protein sequencer using the manufacturer's standard program NORMAL-1. The resulting phenylthiohydantoin amino acid fractions were subsequently identified using an on-line (ABI) HPLC Model 120A. Search of the NBRF-PIR protein sequence database, National Biomedical Research Foundation, Division of Research Resources National Institute of Health, Bethesda, MD, ID# A20595 (1989) revealed that this protein differs from a collagenase inhibitor isolated from human amniotic fluid (which itself appears to be identical to the fibroblast inhibitor from human skin except for one different residue) by only two amino acids at position 17 (valine instead of leucine) and at position 27 (alanine instead of proline) for 28 N-terminal residues as described by Docherty et al., Nature 318, 66-69 (1985). The samples were not reduced and alkylated, so (---) represents bonafide blanks adjusted for homology to expected cysteine residues of previously described collagenase inhibitors. All three residues are, on the basis of Docherty et al., (1985); Carmichael et al., Proc. Natl. Acad. Sci. USA 83, 2407-2411 (1986). Sequence differences between this collagenase inhibitor and others previously described, for example by Murray et al., J. Biol. Chem. 261, 4154-4159 (1986), may be attributed to species variations and/or variable forms of this protein in the same tissue, although CDI appears as a single band on SDS-PAGE analysis followed by silver staining, the apparent Mr of which was determined to be 27,650.

The N-terminal sequence of CDI is shown in Table II. Table II: NH2-terminal protein sequence of the bone-derived inhibitor

Other collagenase inhibitors can also be used. It is possible to use a material that is simply a peptide that is complementary to and binds to the active site of collagenase, or that inhibits the utilization of divalent ions by collagenase, or an antibody that blocks collagenase activity, and compounds that inhibit the activation of precursor collagenase to active collagenase, either directly or by inhibiting the enzymes required to activate collagenase. Due to their specificity, these compounds are not cytotoxic. Other proteins include platelet factor 4 and tissue inhibitory metalloproteinase (TIMP), both of which are collagenase inhibitors and could also contain a peptide sequence that has anti-collagenase and anti-angiogenic activity.

The following demonstrates the discovery of the additional activities that characterize CDI and the effective dose of CDI to inhibit angiogenesis, as compared to inhibition of collagenase activity in vitro. CDI was purified from cartilage as described above.

Example 1: Inhibition of the proliferation of capillary endothelial cells by the CDI

The first in vitro studies conducted with the CDI focused on the effects of this factor on one of the important components of angiogenesis, namely the proliferation of capillary endothelial cells (BCE).

A. Inhibition of endothelial cell proliferation, determined by colorimetric determination of the activity of cellular acid phosphatase and by electronic cell counting.

The colorimetric determination of cellular acid phosphatase activity, described by Connolly et al., J. Anal. Biochem. 152, 136-140 (1986), is used for rapid and sensitive screening of small amounts of fractions from whole columns in the purification scheme for the Inhibition of the proliferation of capillary endothelial cells and is based on color measurement.

By screening each of the successive column steps using this assay, it was found that two different biological activities, anti-collagenase and anti-proliferative activity, were co-purified during the purification regimen. The same results were further obtained by using a second assay to examine BCE proliferation by electronic cell counting. In these studies, fractions enriched for collagenase inhibitory activity after each chromatographic step in the purification regimen were tested for their ability to inhibit acidic fibroblast growth factor (aFGF) stimulation of BCE proliferation. CDI was found to significantly inhibit this proliferation in a dose-dependent manner. Purified CDI inhibited aFGF-stimulated proliferation of capillary ECs with an IC₅₀ of 1.5 mg/kg. (the concentration of inhibitor at which 50% inhibition of proliferation is observed) from 50 to 75 nM. A comparison of the IC50 values (the inhibitor concentration at which 50% inhibition of growth factor-induced stimulation is observed) of the CDI from four consecutive steps determined by cell counting assays shows an increase in the strength of its anti-proliferative activity as the collagenase inhibitor is purified. IC50 values: A-1.5m activity: 14 µg/ml, Bio-Rex-70 activity: 9 µg/ml, Sephadex-G-75 activity: 4 µg/ml, reversed phase HPLC activity: 0.8 µg/ml.

B: Inhibition of endothelial cell proliferation, determined based on DNA synthesis.

In studies to determine the effect of CDI on DNA synthesis by capillary ECs, CDI suppressed radioactive thymidine incorporation by capillary ECs in response to aFGF in a dose-dependent manner with an IC50 of 300 nM.

C: Comparison of the inhibition of endothelial cell proliferation by CDI with controls.

To investigate the specificity of the anti-proliferative effect of the inhibitor, a number of other substances that are particularly important as control factors were tested for their ability to inhibit the proliferation of capillary ECs. Other enzyme inhibitors, e.g. egg trysin inhibitor, pancreatic trypsin inhibitor and α2-macroglobulin, a relatively nonspecific collagenase inhibitor, as well as chondroitin sulfate A, a glycosaminoglycan found in cartilage, had no significant effect (less than 20% inhibition or stimulation at all doses tested) on growth factor-stimulated capillary EC proliferation, even at concentrations of 50 µg/ml. Furthermore, these factors do not inhibit angiogenesis in vivo. In addition, although the inhibitor inhibits aortic endothelial cells (AF) at an effective concentration that is much higher (IC50 = 40 µg/ml) than for capillary endothelial cells (IC50 = 14 µg/ml), the inhibitor has no detectable inhibitory effect on the growth of non-endothelial cell types, such as bovine aortic smooth muscle cells and Balb/c 3T3 cells.

Example 2: Inhibition of migration of capillary endothelial cells by the CDI

Another important component of angiogenesis is the migration of capillary ECs in response to an angiogenic stimulus. See the review by Ausprunk et al., J. Microvasc. Res. 14, 53-65 (1977).

During purification, the EC migration inhibitor was co-purified with the collagenase inhibitor. The IC50 values were 143, 38 and 8 µg/ml for the CDI obtained by A-1.5m, Bio-Rex-70 and Sephadex G75 chromatography, respectively. The purified CDI inhibits capillary EC migration with an IC50 of 96 nM.

Example 3: Inhibition of neovascularization by the CDI in vivo

The chicken chorioallantoic membrane assay (CAM) described by Taylor and Folkman, Nature (London) 297, 307-312 (1982) was used as a third bioassay to determine whether CDI can inhibit neovascularization in vivo. The effect of CDI on growing embryonic vessels was studied using chicken embryos in which capillaries appear in the yolk sac after 48 h and grow rapidly over the next 6 to 8 days.

On the third day of development, the fertilized chicken embryos were removed from their shells and placed in plastic Petri dishes (1005, Falcon) and kept at 37ºC in a humidified atmosphere with 5% CO₂. On the sixth day, 4 µg samples of the purified CDI were mixed in methylcellulose disks and applied to the surfaces of the growing CAMs above the dense subectodermal plexus. After CAMs had been exposed to CDI for 48 hours, the avascular zones, devoid of capillaries and small vessels, were examined using a binocular dissecting microscope at x7-10 magnification. Then, the CAMs were intravascularly injected with Indian ink/liposyn as previously described by Taylor and Folkman (1982).

The results show significant inhibition, as evidenced by a large avascular zone induced by 4 µg of purified CDI (14.5 µM) applied locally in methylcellulose disks to the normally developing chorioallantoic membrane within 48 h of inhibitor exposure. In contrast, control CAMs implanted with empty methylcellulose disks never developed avascular zones. Histological examination of the CDI-treated CAMs revealed a mesoderm that was thinner than normal and avascular, compared to controls that showed normal vascular development.

The centers of the avascular zones were completely devoid of Indian ink-filled capillaries, with the centers of the zones containing the methylcellulose discs. This effect was observed in 33% of eggs examined in at least two separate sets of CAM assays using several different lots of cartilage starting material. Histological sections of CAMs after 8 days (x 800) showed normal capillary development in the untreated controls and controls containing an empty methylcellulose disc. In contrast, capillaries were absent in the experimental CAMs treated with 4 µg of CDI. The anti-angiogenic activity of CDI was determined using this CAM assay. Assays as it was purified. Dose-dilution curves were constructed for the inhibitor from each of the successive purification steps to homogeneity. Purified CDI causes 100% avascular zones in the CAM at 4 µg, 25 µg of CDI from the G-75 chromatography step causes 100% avascular zones in the CAM, 50 µg of CDI from the Bio-Rex 70 column causes 80% avascular zones, and CDI from the A-1.5 m step results in no inhibition of vascularization in the CAM.

The cartilage-derived inhibitor is a potent inhibitor of angiogenesis, as inhibition is observed at up to 4 µg. In comparison, the lowest reported doses for previously described inhibitors of angiogenesis tested by the CAM assay alone are 50 µg protamine (which is also cytotoxic at this concentration) as described by Taylor and Folkman (1982), 200 µg bovine vitreous extract as reported by Lutty et al., Invest. Opthalmol. Vis. Sci. 24, 53-56 (1983), and 10 µg platelet factor 4 as reported by Taylor and Folkman (1982). In comparison, the cartilage-derived inhibitor is an extremely potent inhibitor of angiogenesis in vivo. The lowest reported doses of angiogenesis inhibitors acting as combinations are, for example, heparin (50 µg) and hydrocortisone (60 µg) and B-cyclodextrin tetradecasulfate (14 µg) and hydrocortisone (60 µg), as reported by Eisenstein et al., Am. J. Pathol. 81, 1-9 (1987).

Example 4: Inhibition of capillary tube formation by the CDI

Using the directed migration assay of capillary endothelial cells from cultured BCE cells into collagen gels and by determining whether capillary-like tubes are formed by cultured BCE cells, it was shown that CDI is an effective inhibitor of angiogenesis. In the control collagen gels that were not treated with the inhibitor, a monolayer of BCE cells was formed in the form of a network of branched, anastomosing strands of endothelial cells that resemble capillary beds in vivo. In contrast, collagen gels treated with CDI show a severe disruption of the reorganization of the BCE cells, so that organization into capillary-like tubes appears to be prevented.

Example 5: Inhibition of tumor-induced angiogenesis using the rabbit corneal ash assay

In the above examples, CDI purified from bovine scapular cartilage or scapular chondrocyte conditioned media was shown to inhibit growth factor-stimulated capillary endothelial cell proliferation and migration in vitro at nanomolar concentrations. Since tumor growth is dependent on neovascularization, it was investigated whether CDI suppresses tumor-induced angiogenesis using the rabbit corneal ash assay. V2 carcinoma implants were used to generate the neovascular response in the normal avascular cornea. Sustained-release polymer pellets of ethylene-vinyl acetate copolymer (Elvax) were each impregnated with 40 µg of CDI and implanted between the tumor and the corneal margin.

The tumors induced the growth of vessels 4 to 6 days after implantation. By determining the mean maximum vessel lengths, it was determined that the CDI significantly inhibited the growth of capillary blood vessels in the treated eyes compared to untreated eyes (45 to 64% inhibition).

Example 6: Preparation of pharmaceutical compositions containing CDI as an active ingredient

Examples 1 to 3 demonstrate that a single tissue-derived macromolecule, CDI, is a potent inhibitor of angiogenesis. In vitro studies using this inhibitor demonstrate that it negatively modulates at least three major components of angiogenesis, namely, capillary endothelial cell proliferation and migration and tube formation. The CAM studies demonstrate that it further inhibits angiogenesis in an in vivo model.

Similar results were obtained with the chondrocyte-derived inhibitor, which was tested for inhibition of capillary endothelial cell proliferation and migration. When comparing the inhibition of angiogenesis, the chondrocyte-derived material showed higher activity than the cartilage-derived CDI, which is not glycosylated after isolation using acid precipitation. The chondrocyte material is glycosylated. Materials that are naturally expressed and glycosylated are believed to have higher activity than materials that are deglycosylated or expressed from recombinant sequences. However, they can all be used in a suitable pharmaceutical carrier. for pharmaceutical use.

The pharmaceutical compositions are prepared using CDI as the active ingredient to inhibit angiogenesis for the specific application. The application can be either topical or localized. For topical application, the purified CDI is combined with a carrier to deliver an effective dose based on the desired activity (i.e., the effective dose is, for example, 14.5 µM or 4 µg to prevent local angiogenesis and to inhibit capillary endothelial cell migration and up to 100 µM to inhibit capillary endothelial cell proliferation). For the treatment of diseases such as psoriasis, a topical CDI composition is applied to the skin. The carrier may be in the form of an ointment, cream, gel, paste, foam, aerosol, suppository, compress or gel stick.

A topical CDI composition for treating some of the ocular conditions discussed above consists of an effective amount of CDI in an acceptable ophthalmic vehicle, e.g., buffered saline, mineral oil, vegetable oils such as corn or arachis oil, petrolatum, Miglycol 182, alcohol solutions, or liposomes or liposome-like products. Any of these compositions may further contain preservatives, antioxidants, antibiotics, immunosuppressants, and other biologically or pharmaceutically active agents that do not adversely affect the CDI.

The CDI compositions for local or regional administration, for example into a tumor, contain generally an inert diluent. Solutions or suspensions used for parenteral, intradermal, subcutaneous or topical administration may contain the following components: a sterile diluent such as water for injection, saline, fixed oils, polyethylene glycols, glycerin, propylene glycol or other synthetic solvents, antibacterial agents such as benzyl alcohol or methylparabens, anti-oxidants such as ascorbic acid or sodium bisulfite, complexing agents such as ethylenediaminetetraacetic acid, buffers such as acetates, citrates or phosphates and tonic agents such as sodium chloride or dextrose. The parenteral preparation may be in ampoules, disposable syringes or multiple dose containers made of glass or plastic.

For directed internal topical applications, for example, for the treatment of ulcers or hemorrhoids, or other lesions of mucous membranes, the CDI composition may be in the form of tablets or capsules which may contain any of the following ingredients, or compounds of a similar nature: a binder such as microcrystalline cellulose, gum tragacanth or gelatin, a carrier such as starch or lactose, a disintegrant such as alginic acid, primogel or corn starch, a lubricant such as magnesium stearate or sterote, or a glidant such as colloidal silica. When the dosage unit form is a capsule, it may contain, in addition to the material of the above type, a liquid carrier such as a fatty oil. In addition, the dosage unit forms may contain various other materials which modify the physical form of the dosage unit, for example, coatings of sugar, shellac or other enteric agents.

In a preferred form, the CDI is administered in combination with a biodegradable, biocompatible polymeric implant that releases the CDI at a selected site for a controlled period of time. Examples of preferred polymeric materials are polyanhydrides, polyorthoesters, polyglycolic acid, polylactic acid, polyethylene vinyl acetate, and copolymers and mixtures thereof. The CDI may also be administered locally by an infusion pump, for example of the type used to deliver insulin or chemotherapeutic agents to specific organs or tumors.

Claims (4)

1. Isolated and purified protein, wherein the protein: a) is a normal component of cartilage, (b) has an apparent molecular weight as determined by SDS-PAGE under reducing conditions of about 34 kD when isolated without exposure to acidic pH or of about 28 kD after exposure to pH 2.0, c) is a collagenase inhibitor, and d) produces an avascular zone in a chorioallantoic membrane sample when used in the amount of about 4 µg. 2. A pharmaceutical composition comprising the isolated and purified protein of claim 1 and a pharmaceutically acceptable carrier. 3. Protein according to claim 1 or composition according to claim 2 for use in medicine. 4. Use of a protein according to claim 1 or a composition according to claim 2 in the manufacture of a medicament for inhibiting neovascularization.